The paper introduces ResUNet++, an advanced architecture for medical image segmentation, specifically designed for polyp detection and segmentation during colonoscopy examinations. The goal is to improve the accuracy of computer-aided detection (CAD) systems, which can help endoscopists resect abnormal tissue and reduce the risk of polyps developing into cancer. ResUNet++ is an improved version of the ResUNet architecture, incorporating residual blocks, squeeze and excitation blocks, Atrous Spatial Pyramidal Pooling (ASPP), and attention blocks. Experimental evaluations on the Kvasir-SEG and CVC-612 datasets show that ResUNet++ outperforms the state-of-the-art U-Net and ResUNet architectures, achieving higher dice coefficients and mean Intersection over Union (mIoU) scores. The paper also discusses the contributions, related work, and implementation details, highlighting the effectiveness of the proposed architecture in producing more accurate and semantically meaningful segmentation results.The paper introduces ResUNet++, an advanced architecture for medical image segmentation, specifically designed for polyp detection and segmentation during colonoscopy examinations. The goal is to improve the accuracy of computer-aided detection (CAD) systems, which can help endoscopists resect abnormal tissue and reduce the risk of polyps developing into cancer. ResUNet++ is an improved version of the ResUNet architecture, incorporating residual blocks, squeeze and excitation blocks, Atrous Spatial Pyramidal Pooling (ASPP), and attention blocks. Experimental evaluations on the Kvasir-SEG and CVC-612 datasets show that ResUNet++ outperforms the state-of-the-art U-Net and ResUNet architectures, achieving higher dice coefficients and mean Intersection over Union (mIoU) scores. The paper also discusses the contributions, related work, and implementation details, highlighting the effectiveness of the proposed architecture in producing more accurate and semantically meaningful segmentation results.